| --- |
| license: mit |
| tags: |
| - healthcare |
| - machine-learning |
| - lightgbm |
| - stroke-prediction |
| - binary-classification |
| --- |
| |
| # Stroke Risk Predictor - LightGBM |
|
|
| ## Overview |
|
|
| This model predicts stroke risk using healthcare-related patient features. |
|
|
| ## Model Type |
|
|
| LightGBM Classifier |
|
|
| ## Inputs |
|
|
| - Age |
| - Hypertension |
| - Heart Disease |
| - Average Glucose Level |
| - BMI |
| - Smoking Status |
| - Other health indicators |
|
|
| ## Output |
|
|
| - 0 → No Stroke Risk |
| - 1 → Stroke Risk |
|
|
| ## Performance |
|
|
| Accuracy: 91.10% |
|
|
| Recall: 28% |
|
|
| ROC-AUC: 78.74% |
|
|
| ## Usage |
|
|
| ```python |
| import joblib |
| |
| model = joblib.load("lightgbm_model.joblib") |
| |
| prediction = model.predict(data) |
| ``` |
|
|
| ## Author |
|
|
| Mithun |
